利用广义形态滤波与差分熵的电能质量扰动检测

Detection of Power Quality Disturbances Utilizing Generalized Morphological Filter and Difference-entropy

数学形态学是一种非线性信号处理方法,在电力信号的分析上具有独特的优势,而信息熵是一种典型的无序性测量指标,在电力系统中也具有巨大的应用潜力。在研究数学形态学算法及信息熵基础理论的基础上,提出一种基于广义形态滤波与差分熵的电能质量扰动信号检测方案,该算法利用广义形态滤波器作为前置滤波单元对信号进行预处理,再利用差分熵度量信号的复杂程度,获取扰动信号的特征量,从而实现扰动的定位。对6种理想信号进行了仿真,结果表明,该算法对非过零时刻发生扰动的信号具有较好的适应性,受噪声强度及信号幅值突变幅度影响较小。

As a processing method on non-liner signals, mathematical morphologic has inimitable advantages on electric signals analysis. Moreover, Shannon entropy is a typical measure index on disorder and its application in power system has a tremendous potential. Researching on the theories of mathematical morphologic and Shannon entropy, a new detection method on basis of generalized morphologic filter and difference-entropy was proposed. Signals were processed by generalized morphologic filter which was used for pre-positive filter unit, then, to obtain eigenvalue for disturbances location, complexity of signals was measured by difference-entropy. A simulation on six ideal signals is done and its results show that this method has good adaptability to signals whose disturbance doesn＇t occur at cross-zero point, furthermore, noise intensity and extent of magnitude variety has little impact on this proposed method.